Deep‐learning based on‐chip rapid spectral imaging with high spatial resolution
Jiawei Yang,
Kaiyu Cui,
Yidong Huang,
Wei Zhang,
Xue Feng,
Fang Liu
Affiliations
Jiawei Yang
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China; Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University, Beijing 100084, China
Kaiyu Cui
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China; Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University, Beijing 100084, China; Corresponding authors.
Yidong Huang
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China; Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University, Beijing 100084, China; Bejing Academy of Quantum Information Science, Beijing 100084, China; Corresponding authors.
Wei Zhang
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China; Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University, Beijing 100084, China; Bejing Academy of Quantum Information Science, Beijing 100084, China
Xue Feng
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China; Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University, Beijing 100084, China
Fang Liu
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China; Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University, Beijing 100084, China
ABSTRACT: Spectral imaging extends the concept of traditional color cameras to capture images across multiple spectral channels and has broad application prospects. Conventional spectral cameras based on scanning methods suffer from the drawbacks of low acquisition speed and large volume. On-chip computational spectral imaging based on metasurface filters provides a promising scheme for portable applications, but endures long computation time due to point-by-point iterative spectral reconstruction and mosaic effect in the reconstructed spectral images. In this study, on-chip rapid spectral imaging was demonstrated, which eliminated the mosaic effect in the spectral image by deep-learning-based spectral data cube reconstruction. The experimental results show that 4 orders of magnitude faster than the iterative spectral reconstruction were achieved, and the fidelity of the spectral reconstruction for the standard color plate was over 99% for a standard color board. In particular, video-rate spectral imaging was demonstrated for moving objects and outdoor driving scenes with good performance for recognizing metamerism, where the concolorous sky and white cars can be distinguished via their spectra, showing great potential for autonomous driving and other practical applications in the field of intelligent perception.
